Morphological Traits and Nutrient Uptake of Dragon’s Head (Lallemantia Iberica Fish.): Effects of Sowing Seasons and Different Fertilizer Sources

The effects of sowing seasons and chemical, organic and biological fertilizer sources were explored on morphological traits of dragon’s head (Lallemantia iberica Fish.) in a eld experiment based on a randomized complete block design in three replications and six treatments. The fertilization treatments included organic fertilizers (vermicompost, manure, and humic acid), biological fertilizer (Thiobacillus mixed with sulfur), chemical fertilizer (macro NPK), and control (no-fertilization). The recorded traits included leaf area index, plant height, number of ower cycles per plant, number of achenes per plant, number of seeds per plant, 1000-seed weight, number of auxiliary branches, plant diameter, and the uptake of nitrogen, potassium, phosphorus, calcium, and sodium. The results showed that the winter sowing outperformed the spring sowing by a wide margin and results obtained for morphological traits were signicantly higher than those of the spring sowing. The fertilization of the plants in both sowing seasons, especially in the winter sowing, improved yield and yield components. The uptake of N and K was not inuenced by the sowing season and fertilizer type, whereas the uptake of P, Ca, and Na was inuenced by these factors but with a slight difference. It can, so, be inferred that the uptake of nutrients in dragon’s heads is less inuenced by the environment. Overall results indicated that the improvement of morphological traits in the ecological conditions of Azerbaijan region (Iran) was most notable for winter sowing of dragon’s heads with the use of Thiobacillus and vermicompost fertilizers.


Introduction
The safety of crops produced by different agronomic systems in the world in terms of the residues of pesticides and agrochemicals and their harmful effects on the health of humans and the environment has drawn attention to production methods and inputs used in crop production. Following the recent crises of environmental pollution, extensive attempts have been made to nd solutions to improve the safety of soil and crop quality, remove pollutants, and protect the sustainability of natural ecosystems 1 . Organic fertilizers constitute a major pillar of soil fertility as they have bene cial impacts on the physical, chemical and biological properties and the fertility of soils. These fertilizers increase soil organic matter and improve its fertility in the light of their impact on improving the chemical characteristics of soils including acidity, cation exchange capacity, the activity of microorganisms, and nutrient availability 2 .
Biological fertilizers, which are sometimes used as an alternative to chemical fertilizers and the other times as their supplement, can guarantee the sustainability of agricultural systems 3 . The global approach in medicinal plant production is now towards improving the quantity, quality, and health of the active substances. So, the feeding of these plants with organic and biological fertilizers seems to be best suited with the goals of their production and can improve their quantitative and qualitative yields 4,5 .
The history of using medicinal and aromatic plants to treat humans is as long as the history of mankind.
Although the application of chemical and synthetic drugs has been expanded enormously in the last halfcentury, its harmful impacts on human life have revived interests in medicinal plants 6 .

Page 3/14
The plant species of the family Lamiaceae have traditionally been used as medicinal plants 7 . The medicinal and aromatic plants of this family are regarded as a rich genetic reserve because of their ecological resilience and extensive distribution in diverse climates. They have also extensive applications in the cosmetic and health industries due to their various aromatic compounds 8 . Dragon's head (Lallemantia iberica Fish.) is a plant species from this family that has a lot of applications. All parts of this plant species, including its leaves and seeds, can be consumed economically. Some applications of dragon's head include the extraction of oil from the seeds, the production of mucilage from the seeds, the extraction of essential oil from the vegetative parts, the application of seed meal (after oil extraction) as food and feed, the use of leaves and green twigs before owering as edible vegetable, and its application as green fertilizer. Due to their mucilage compounds, the seeds of dragon's heads are good for cough 9 .
These seeds are also used to treat neural, hepatic, and renal disorders and have traditionally been consumed as laxative, forti er, sexual arouser, diuretic, and expectorant 10,11 .
Dragon's head is still grown in different parts of Iran sparsely, including provinces of Alborz, Guilan, Qazvin, Azerbaijan, Kurdistan, Kermanshah, Hamedan, Lorestan, Isfahan, Chaharmahal and Bakhtiari, Fars, Semnan, and Tehran 12 . A few studies have dealt with this plant species in the world in general, and in Iran in particular, but there are indications that it should be planted by the conventional patterns without the excessive use of chemical fertilizers 13 . The most recent research revealed that the appropriate combination of nano fertilizers enhanced the quality of yield components and antioxidant properties of dragon's head 14 .
Continuing our interest for the improvement of qualitative and quantitative traits of dragon's head, the present study aimed to explore the effect of the sowing seasons and to explore the best fertilization system for the improvement of morphological traits and nutrient uptake of this medicinal plant with an emphasis on its organic production.

Experiments design
The study was carried out in the research farm of Faculty of Agriculture, Urmia University in Western Azerbaijan Province, Iran (longitude 45°10' east, latitude 37°44' north, and elevation 1338 meters from sea level) in the 2017-2018 growing season. The properties of the soil were as presented in our previous work 14 . The physicochemical properties of the applied organic fertilizers are shown in Table 1. The fertilizer requirement was calculated according to the results of soil analysis, fertilizer recommendations, and the N content of the applied organic fertilizer. Then, they were incorporated with the soil before sowing.
The study was a factorial experiment laid upon a randomized complete block design with three replications. The factors included sowing season and fertilizer source. The sowing season was assigned to the rst factor including two levels of winter sowing and spring sowing. The fertilizer source was assigned to the second factor at ve levels of manure, chemical NPK fertilizer, vermicompost, humic acid, and Thiobacillus + sulfur. The seeds were collected from a local landrace in the south of Western Azerbaijan Province. The blocks were developed on the farm on November 27, 2017. After an autumn plowing and land leveling, the sowing rows were created. The experimental plots were set at 6 m 2 . The rst experiment was set at the winter, and sowing was performed on November 28, 2017 by the row sowing technique. On-row spacing was set at 1 cm and between-row spacing at 25 cm. The second experiment was set at spring, and sowing was conducted on February 27, 2018. The thinning, gap-lling and weeding operations were performed conventionally during the growing season. To determine the morphological traits at the harvest time, 10 plants were randomly harvested from each plot to assess the traits. To estimate the yield, two side rows and 0.5 m from both ends of the plots were eliminated as the marginal effect. To measure the leaf area, three plants were randomly harvested from the second and third rows after eliminating 1 m from their ends. The leaf area of the three plants was rst determined by the copying technique and the leaf area index was nally yield for different treatments.

Determination of morphological traits and nutrients
To measure the plant height, 10 plants were randomly taken from an area of 1 m 2 at each plot. Then, their height was recorded with a ruler in cm and it was averaged over each plot to yield the plant height. To measure the plant diameter, 10 randomly selected plants were harvested. After their roots were detached, the end part of the plants was measured with a caliper to nd out plant diameter and its average was recorded for each plot. The auxiliary branches of 10 random plants in each plot were counted and averaged to nd out the number of auxiliary branches. The number of achenes was counted on 10 plants and its average was recorded as the number of achenes per plant. The number of seeds was also determined for 10 plants and was averaged to yield this trait. The number of owers was counted on 10 plants and was averaged to yield the number of owers per plant. To determine the 1000-seed weight, 500 seeds were counted and their weight was doubled, and this was repeated three times to yield the 1000-seed weight.
Nutrients such as nitrogen (N), potassium (K), phosphorus (P), calcium (Ca), and sodium (Na) were determined as follows: Nitrogen content was measured by sulfuric acid and the Kjeldahl system. Phosphorus content was estimated by the method of dry combustion with HCl. Then, the samples were read by spectrophotometer. To measure potassium content, the samples were placed in a furnace and they were then undergone dry combustion with HCl. Then, a amephotometer (the model Clinical pfp7) was used to read potassium standards rst and then the main samples. To nd out calcium and sodium content of foliage and shoot, 1 g of foliage was ground and screened and was placed in a furnace at 550°C for 24 hours. The samples were adjusted to 100 mL after digestion with dry combustion method (with HCl). Then, a ame photometer (model Clinical pfp7) was used to read the standards rst and then the main samples.
Data were analyzed on the basis of the expected value of the base design and using the SAS 9.2 software package. Means comparison for the traits was carried out by multiple-range test at the P <0.05 level. The graphs were drawn by the MS-Excel software package.

Morphological traits
Leaf area index -Leaf area index (LAI) is a major variable in climatic, ecological and agricultural research. Means comparison revealed that the maximum LAI (5.02) was obtained from the vermicompost-fertilized winter sowing treatment and the minimum (1.89) from the unfertilized spring sowing treatment ( Figure  1). The highest LAI (4.08) among the spring-sown plants was related to those fertilized with Thiobacillus.
The application of biosulfur (containing Thiobacillus), especially to spring-sown plants, seems to be a strategy to increase P availability in soils via oxidizing sulfur and in uencing soil pH, which can improve plant growth 15 . The leaves of dragon's heads are another economic part of this medicinal plant. The leaves and green twigs of the pre-owering plants are consumed as fresh vegetables or as green manure. The expansion of leaf area is important in the sense that it develops plant area and allows the interception of more radiation, and subsequently, this process increases photosynthesis and assimilation, resulting in higher growth rate. The fertilizers used here provided the plants with various amounts of nutrients, which was responsible for the difference among the treatments in LAI.
lant height -According to the results of means comparison, the winter sowing increased the plant height when compared to the spring sowing (76.6 cm versus 70.7 cm). The highest plant heights were related to the treatments of vermicompost (78.2 cm), Thiobacillus (77.9 cm), and manure (77.9 cm), respectively and the lowest (61.9 cm) was related to the control ( Figure 2). Since nutrient de ciency is a key factor in dictating plant height, it seems that no-fertilizer treatment exhibited lower growth because of nutrient de ciency whereas all fertilization treatments had a positive effect on the vegetative growth of the plants. and humic acid treatment (65 achenes). However, the spring-sown plants fertilized with Thiobacillus (109 achenes) and vermicompost (104 achenes) exhibited more achenes than their control counterparts, but even these higher number of achenes could not compete with that of the unfertilized winter-sown plants, which produced 143 achenes (Figure 4). In a study on the effect of high-input (urea and triple superphosphate) and low-input (Nitroxin, Barvar-2 phosphate, and biosulfate) fertilization system along with the effect of irrigation frequency on the morphological traits of dragon's heads, it was reported that these factors affected the number of achenes signi cantly and the high-input system produced more achenes 20,21 (34 achenes), also, reported that in the monocropping system of dragon's heads, the biofertilizer-treated plants produced more achenes per plant than the unfertilized plants. An effective way to enhance rainfall use e ciency is to make spring rainfalls coincide with the initiation of vegetative and reproductive growth by selecting a proper sowing date 22 .
Number of seeds per plant -The simple and interactive effect of the fertilizer source and sowing season was signi cant (p < 0.01) on the number of seeds per plant (Table 4). According to means comparison ( Figure 5), the fertilizer treatments did not differ to one another signi cantly in the spring sowing, and their seed number per plant (85 seeds) was the lowest. But, these treatments in the winter sowing differed signi cantly. The highest number of seeds per plant in the winter sowing was obtained from the NPK fertilizer (181 seeds) and manure (175 seeds). The next ranks were related to the application of vermicompost (162 seeds) and Thiobacillus (152 seeds) followed by humic acid (131 seeds) and nofertilization treatment (126 seeds). To explain the increase in the seed number per plant 19 , argue that in a treatment in which soil moisture is appropriate, chlorophyll growth and then the number of achenes per plant will increase and this, in turn, will enhance the number of seeds per plant. Consequently, a severe competition will initiate between the lling grains that are strong sinks for photosynthate absorption and thereby smaller light-weight seeds will be produced.
Thousand-seed weight -It is evident that the highest 1000-seed weights were related to the application of Thiobacillus or vermicompost to the winter-sown plants (5.68 and 5.66 g, respectively) and the lowest was 2.31 g observed in the unfertilized spring-sown plants ( Figure 6). Also, the application of Thiobacillus to the spring-sown plants exhibited a 1000-seed weight of as high as 4.48 g. In the winter sowing, the treatment of NPK fertilizer (3.85 g) had lower 1000-seed weight than the control (4.21 g), but their difference was insigni cant. This is related to the fact that the application of NPK fertilizer produced the highest number of seeds per plant and it is obvious that when the number of seeds increases, their weight decreases due to competition. Based on the results, the higher seed yield was probably due to the positive effect of vermicompost and winter sowing on 1000-seed weight. The application of biofertilizers in the study of 21 enhanced the yield and yield components (e.g. 1000-seed weight, the number of pods per plant, the number of seeds per pod, the number of auxiliary branches, and plant height) of dragon's heads and peas so that they recommended the application of biofertilizers in the intercropping system of this plant. A sowing date that has recently been considered is winter sowing in which the sowing date is of crucial importance because the seeds should be sown when the air and soil are cold enough to prevent germination. Winter sowing aims at maximizing the use of rainfall, especially in regions where no land preparation is possible in springs due to rainfalls 23 . Figure 7, the winter-sown plants showed the highest number of auxiliary branches when they were fertilized with vermicompost (6.9 branches) or Thiobacillus (6.4 branches). In contrast, the spring-sown plants exhibited the lowest number of auxiliary branches when they were not fertilized (2.4 branches) or were fertilized with humic acid (2.9 branches). The highest number of auxiliary branches in the spring-sown plants was related to those treated with Thiobacillus (5.6 branches). The signi cant difference of the fertilization treatments with the control in both seasons implies that the application of fertilizer facilitates nutrient access for plants and help them establish better. In a study on the effect of time on some traits of dragon's heads 24 , concluded that the number of auxiliary branches was signi cantly higher in the early sowing than in the late sowing. Our results about the effect of fertilization on the number of auxiliary branches of the dragon's heads are consistent with the results of 25,26, for Silybum marianum L., for fennel, and 27 for lentil, who have all expressed the signi cant effect of biofertilizers on the number of main and auxiliary branches of the studied plant species.

Number of auxiliary branches -As depicted in
Plant diameter -The highest plant diameter was 9.0 cm observed in the winter-sown plants that were treated with vermicompost. The next highest plant diameters were related the NPK-fertilized winter-sown plants (8.05 cm), manure-fertilized winter-sown plants (7.89 cm), and Thiobacillus-treated winter-sown plants (7.88 cm), which were all ranked in the same statistical group. The application of humic acid and no-fertilizer to the winter sowing resulted in the lowest plant diameters of 6.79 and 6.28 cm, respectively. All fertilization treatments in the spring sowing showed the minimum stem diameter and were ranked in the same statistical group (Figure 8). Since stems are a sink of assimilates, the improvement of nutritional conditions by consuming fertilizers increased assimilates, resulting in the enhancement of stem diameter. The failure of humic matter in improving plant growth has been reported in several research works. For example, a study on oregano reported that the plants treated with humic matter produced a stem diameter even lower than the control 28 . In their investigation of the effect of biofertilizers on the quantitative yield and morphological traits of dragon's heads exposed to severe water stress 29 , recommended the application of Barvar-2 phosphate biofertilizer under optimal irrigation conditions to achieve the highest number of pods per plant, seeds per plant, seed yield, biological yield, and stem diameter.

Nutrient uptake
Despite the insigni cant effect of the fertilizer on N content, means comparison by Duncan's multiple range test grouped the treatments as depicted in Figure 9. Among the studied six fertilization treatments, the application of manure and vermicompost produced the highest and no-fertilization treatment produced the lowest N contents (2.67%, 2.53%, and 2.12%, respectively). The remaining treatments were in between these two extremes and did not in uence N content considerably. In terms of phosphorus (P) content, the simple effect of the sowing season and fertilizer type was signi cant (p < 0.01), but their interaction was insigni cant (Table 5). Based on the results of means comparison for this nutrient, the winter sowing increased P content versus the spring sowing (0.73% versus 0.60%). Also, among the fertilization treatments, Thiobacillus was related to the highest P content of 0.76% and the other fertilizers did not differ from no-fertilizer application (0.60%) signi cantly, so they were all placed in the same statistical group. Therefore, in spite of the signi cant impact, there was not a remarkable difference between the fertilization treatments ( Figure 10).
Means comparison for the simple effects ( Figure 11) showed higher Ca content in the winter sowing (7.12%) than in the spring sowing (6.54%). Also, the treatment of humic acid (7.95%) and manure (7.04%) were related to the highest and the no-fertilizer treatment (6.03%) was related to the lowest Ca content. The winter sowing had a slightly but signi cantly higher Na content than the spring sowing (4.74% versus 4.33%). Furthermore, the application of manure was related to the highest Na content (7.06%), but the lowest was related to the no-fertilizer treatment (4.01%). The other fertilizers had almost as high Na content as the control (Figure 12).
It was observed that the application of manure increased the uptake of nutrients. Manure increases soil cation exchange capacity, soil nutrient contents, nutrient availability to plants, nitrogen balance, soil organic matter and humus content, and soil granulation, thereby increasing its porosity and improving its structure 30 . For these reasons, the growth and expansion of root systems are increased in soils treated with organic fertilizers where plants can grow and absorb nutrients well under the appropriate physical and chemical conditions created by the organic fertilizers 31 . Likewise 32 , reported for coriander that the application of different fertilizers including manure enhanced the uptake of N, P, K, and Na signi cantly as compared to the no-fertilizer treatment. Although the phenotypic representation of traits is in uenced by genetics, environment, and their interaction 33 , these in uences may vary with plant and trait.

The analysis of variance of morphological and yield traits
The results of the analysis of variance (ANOVA) indicated that the effects of fertilization and sowing date were signi cant on leaf area index (LAI) and plant height (P <0.01), whereas the interactive effect of these factors was signi cant (P <0.01) on LAI but insigni cant for the plant height of the dragon's heads ( Table   2).
The interactive effects of the sowing season and fertilizer were signi cant (P <0.01) on the number of ower cycles per plant, number of achenes per plant, number of seeds per plant, 1000-seed weight, number of auxiliary branches and the plant diameter ( Table 3). The results of ANOVA for the number of ower cycles per plant, the number of achenes per plant, number of seeds per plant, 1000-seed weight, and plant diameter indicated that the effects of the sowing season and fertilizer were signi cant (P <0.01). Additionally, the interaction of studied effects was signi cant on the number of auxiliary branches at P <0.05 (Table 3).
According to the results of ANOVA it is evident that the effects of the sowing season and fertilizer and their interaction were statistically insigni cant for the uptake of N and K, whereas both factors were signi cant to phosphorus uptake (P <0.01), unlike their interaction ( Table 4). The data on Ca content revealed that this trait was in uenced by the sowing season at the P <0.05 level and by the fertilizer type at the P <0.01 level, but their interaction was insigni cant for this trait. The two sowing seasons and six fertilizer treatments differed signi cantly (P <0.05) in Na content. Like the other studied elements, the interaction of sowing season × fertilizer was insigni cant for this trait. The insigni cant effect of sowing season and fertilizer on the uptake of some elements means that these traits were not in uenced by sowing season and fertilizer type and there was not a speci c trend in the variations of this trait. Even in elements in which the impacts were signi cant, the differences were not considerable among the treatments. It can, thus, be concluded that the uptake of nutrients is affected by sowing season and fertilizer to a lesser extent.

Conclusion
The obtained results revealed that the winter sowing of dragon's heads outperformed the spring sowing by a wide margin. The fertilization of this plant species in both sowing seasons, especially in winter sowing, improved its morphological traits. Among the different fertilizers, vermicompost and Thiobacillus proved to be more effective whereas the application of humic acid, especially to the spring-sown plants, was the least effective in improving the traits.
The uptake of some elements was not in uenced by the sowing season or fertilizer type. Even the elements that were in uenced by these factors did not exhibit large differences between the treatments. Nonetheless, the highest rate of nutrient uptake was observed in the treatments of manure, humic acid, and Thiobacillus with slight differences. Thus, it can be inferred that the trait of nutrient uptake, especially N and K, is in uenced by the environment and fertilizer to a lesser extent.

Con ict of interest
The authors declare no con icts of interest. ns, *, and ** show insigni cance and signi cance at the P <0.05 and P <0.01 levels, respectively.  ns, *, and ** show insigni cance and signi cance at the P <0.05 and P <0.01 levels, respectively.